ChatGPT adds native image editing with a new, faster visual model

OpenAI has introduced a built-in image editor inside ChatGPT, expanding the product beyond text with a dedicated Images section that allows users to modify photos directly within the interface.

The new feature functions as a standalone visual editing tool powered by an updated image model that OpenAI says operates about four times faster than its predecessor. The editor is designed to handle detailed changes without affecting the rest of an image, addressing a common limitation of earlier generative image systems.

More precise image control

A key improvement is the model’s ability to maintain visual consistency. ChatGPT can now remember how a person or object appeared in previous frames, making it easier to apply multiple edits without introducing distortions or unintended changes.

Users can adjust individual elements of an image — including pose, lighting, clothing, or background — while preserving the overall composition. According to OpenAI, the system also follows prompts more reliably, reducing the trial-and-error approach that often accompanies AI-based image editing.

Availability and significance

The image editor is already available to all ChatGPT users. While the update does not radically change how images are generated, it represents a shift toward more practical, fine-grained control, positioning ChatGPT as a broader creative tool rather than a platform focused primarily on text.

‍